The six sessions π¨ Β· one screen, designed across all of them
The running project is Daybreak's "Manage your subscription" area - the screen where a coffee subscriber pauses, skips, changes their grind or cancels. It is the highest-stakes surface the company owns, because it is the only screen where the customer is actively deciding whether to keep paying, and right now the button that saves the relationship is buried two taps behind the one that ends it.
Before pixels
What the problem actually is, who is hurting, and the one number that would move if you solved it. No screens allowed in this session at all, which is the point of it.
π§ Session 2 Β· easyResearch that changes it
Five interviews, questions that do not lead, and the quotes-to-insights move. Plus the discipline that makes research honest: write down what you expect to find before you go and find it.
ποΈ Session 3 Β· mediumStructure and flow
Information architecture as the cost of nesting, task flows before screens, deliberately ugly wireframes, and the empty-loading-error trio that is the common case on a screen like this one.
π Session 4 Β· mediumThe visual system
Type scale, spacing scale, hierarchy and colour - graded by a live scorer that reads the real rendered pixels and computes the WCAG ratio. Then you meet the anti-lever, and it costs you 42 points.
π¬ Session 5 Β· hands-onInteraction and copy
Feedback within a second, microcopy as interface rather than decoration, forms improved mainly by deletion, a motion budget, and confirmations that name the consequence instead of asking if you are sure.
π€ Session 6 Β· hands-onCritique and handoff
Running a critique that produces decisions instead of opinions, receiving one without defending, the handoff pack that survives you, and writing down the debt you knowingly shipped.
Your passport π«
Clear all three quiz questions in a session and it gets stamped. Six stamps means you have taken one screen from a complaint to a handoff. Stamps live in this browser only.
What this course is not, so you land in the right place π§
| If your question is | Go here instead |
|---|---|
| How do I use AI tools to make design assets faster? | learn-ai-design-with-phoebe |
| How does product management change when drafting is free? | learn-ai-pm-with-phoebe |
| How do I actually build the front end? | learn-html-with-phoebe |
| How do I test whether the change worked, statistically? | learn-experimentation-with-phoebe |
| How do I design something that works for the person using it, and hand it over so it survives? | you are in the right place |
This course is deliberately AI-agnostic. Every judgement in it is one a tool cannot make for you, which is exactly why it keeps its value.
Choose your path πΊοΈ
Three doors into the same conviction: design is a series of decisions you can defend with evidence, not a matter of taste you either have or do not.
What is honest here π§Ύ
The scorer in session 4 is half measurement and half teaching model, and it tells you which is which on its own footer.
| Claim | Status |
|---|---|
| The contrast ratio on the scorer | Really measured. It reads the rendered elements' computed colours, resolves the background, and applies the WCAG 2.1 relative-luminance formula. Change a colour and the number changes because the pixels changed |
| 4.5:1 for body text, 3:1 for large text, 7:1 for AAA | WCAG 2.1 contrast minimums, as specified |
| Hierarchy, spacing, copy, states and findability scores | A teaching model with a stated rubric, calibrated so each trade is visible. Not a measurement of your product |
| "Time to find pause", in seconds | Modelled, not observed. It exists to make nesting cost visible, and the page says so where it appears |
| Daybreak's numbers | A teaching scenario. The company is fictional and every figure about it is part of the exercise |
The knowledge map π§
The whole course at a glance - hover a session to spotlight its concepts, click any node to jump in.